OF WASHINGTON. 29 



With fcecifork Cassididae. 



Without fcecifork. 



Abdomen bent sharply ventrad, labial palpi : Cryptocephalini. 

 of two segments, palpiger distinct, case- { Clythrini. 



bearers Cryptocephalidse ^Chlamydini. 



Abdomen straight. 



Caudal pair of abdominal spiracles on ninth segment, ab 

 dominal tergites and sternites with transverse depres 

 sions, body flattened, maxillary palpi of one or two seg 

 ments Hispidae. 



Spiracles on first eight abdominal segments, no transverse 

 depression (except fold between annulets). 



Prolegs wanting, anus situated on anterior margin of 

 eighth abdominal tergite. bases of eighth abdominal 



spiracles forming horns Donaciidse. 



With anal prolegs. anus notas above. .Chrysomelida;. 

 Lacinia present, anal prolegs single, claw with 



pulvillus Gallerucini. 



Lacinia wanting, prolegs double. 



Labial palpi of two segments, ocelli 



six Chrysomelini. 



Labial palpi of one segment, ocelli 



wanting Eumolpini. 



Labial palpi of one segment, ocelli 

 six Criocerini. 



These five larval types are very distinct. There is less simil 

 arity between some of them than between them and larvae of other 

 families. It seems probable therefore that they indicate a better 

 classification of the Chrysomelidae, ranking it as a superfamily, 

 and dividing it into five distinct families. This is indeed almost 

 the same classification as that of entomologists early in the last 

 century. 



It is obvious that from the many characters which the adult 

 beetles have in common that entomologists have had good reason 

 for considering the Chysomelidas as but one family and its various 

 subdivisions merely as series. But as Dr. Weismann has well 

 shown in his "Studies in the Theory of Descent," treating of 

 lepidopterous larvae, the generic and family relationships are 

 most clearly to be discerned in that stage of insects in which these 

 classes differ most in their habits. He there points out the in 

 congruities between the larvae of Lasiocampa, Clisiocampa, and 

 allied genera, and most of the other genera then included in the 

 Bombycidas. and makes the query whether or not morphological 

 differences do not exist in the adults so that these genera should 

 form a distinct family. Further study of the moths has shown 

 several distinctive characters, notably the wing venation, and the 



